Member
Joined 2009
Paid Member
As you have know, the chasis must be connected to the safety earth of the mains wiring. Try connecting the 'star' ground to the chasis through a 10R power resistor. You may also need to put a pair of diodes in parallel with the resistor, connected opposite way around from each other to protect the resistor from faults. You may also want to try disconnecting the input cable shield from the chasis and connect it to the star ground instead.
Last edited:
On very high frequency band, position of the wires influence capacitance between, so I suppose the channel that have no twisted wires is more sensitive to oscillations. Ground issues are likely too.
I'm a dope. Today I didn't have much time so I rushed in and made new wires to reach around the PS to see if that would help. Being in a hurry I didn't bother with the light bulb and just flipped the switch. One board instantly burned 6 resistors and the other let out a big puff of magic smoke. In my haste, I plugged in the power wires reversed on one of the boards. 
Setting it aside for a few days while I wait for some parts. One step forward, two steps back.
Setting it aside for a few days while I wait for some parts. One step forward, two steps back.
OK I'm back. I ordered new parts and also ordered two new boards. It all arrived yesterday. This morning I replaced the small transistors on the - side of the boards since one of the 2S transistors had popped. That didn't fix it so I just populated the new boards and hooked one of them up. I was missing two 15V zeners so I could only compete one board. Anyway, I brought it up slowly through a light bulb and all looked good. I was able to set the bias and zero the DC offset. All looked fine until I plugged in something to the input and then the bias jumped to 1.2V. So I'm guessing something is oscillating but have no idea how to look for the culprit. I have a scope so if one of you could explain the procedure I would love to go through it and find it. I have one board that behaves and one that doesn't. The only difference between them is that I used Phillips BC556B in this latest build. Is it likely that just using a different brand could be causing this? It is there a way to check with out replacing?
Thanks, Terry
Thanks, Terry
These boards are designed to use 2SC2240/2SA970 for Q1-4 in place of the BC546C/BC556C in the original design. However, the other small transistors are all BC546B/BC556B. I was out of Fairchild BC556B so I used some Phillips models. I think they have a little higher hfE than the fairchilds.
At first glance, considering your observations, Q9 and Q10 keeps attention. I would try to decouple and compensate base to collector by 10 - 100pF, just to see if it helps. Do it as close as possible to device pins.
If such minor parameter difference is critical, there is problem in the design. Something is unstable here, and probably cause earlier damages.
Member
Joined 2009
Paid Member
This amplifier has very high open loop gain and consequently a high unity gain frequency. It will oscillate easily if everything is not right.
My own version, TGM7, oscillates if I take the supply voltages from +/-20V to +/-50V. I found three different kinds of oscillations !
I am revising my pcb design (and wiring) to improve it. It's an interesting learning opportunity to work with a design like this. However, if you just want something to work then it may be best to buy the boards and key parts from Greg, wire it up exactly how he says. His amplifiers have been working just fine for several years in the field - nothing wrong with the design.
My own version, TGM7, oscillates if I take the supply voltages from +/-20V to +/-50V. I found three different kinds of oscillations !
I am revising my pcb design (and wiring) to improve it. It's an interesting learning opportunity to work with a design like this. However, if you just want something to work then it may be best to buy the boards and key parts from Greg, wire it up exactly how he says. His amplifiers have been working just fine for several years in the field - nothing wrong with the design.
This is a brand new board, freshly assembled. I remember now that I had trouble with these transistors when I first assembled my first boards. I wasn't sure because I had the BD139 backwards at first and I had used those Phillips in a DX Super A that wouldn't work until I changed them out so I pulled them out of the SKA as a matter of course. It took reversing the 139 to get it to work. I let this amp play for two or three days and it played fine. It wasn't until I installed it in the case that it started acting up. This new board however won't play nice laying on the bench so I'm hoping it is the transistors. I should know by next week.
Yeah, I don't "just want something to work." I want to learn. Don't think that Greg's design has been trouble free. Reading through the little bit that is not sealed on his forum, there have been many guys who had trouble getting their amp to work and Greg had to send them "fixes". I posted my troubles with this amp on his "public domain" thread and he not only refused to help me, he deleted my posts and any others dealing with them. Evidently, he is still selling PCBs and kits for this amp and wants to discourage others from choosing a different avenue.
Paulo got his working using these boards so I have faith that it is possible. It is a nice sounding amp so I am determined to get it working. Besides, I already built the case.
Blessings, Terry
To be fair, there have been no more issues with people having trouble with the SKA than there would be with any DIY amp, with all the permutations of construction that entails. In particular, I can't remember anyone having persistent issues with oscillation. I've been a member over there since the forum started. I have built two now with no issues at all.
Fixes were only really needed where builders wanted to go outside the standard build parameters. The great thing about Greg is he'd accommodate these people by simming etc. for the new operating conditions or whatever and suggest new parts to enable the amp to work under these conditions.
Strangely, the only real issue was found in really cold conditions where the amp would whine on turn-on, but out excuse here in Australia is it never gets really cold so Greg's test conditions didn't encompass such temperatures. An extra cap fixed this.
He is really helpful even when you ask him anything at all about electronics, including other brands etc. But you can't blame him for not supporting construction of another vendor's boards when they don't work - think about it! Not only that, but the posts he deleted essentially promoted these other boards (that is, made others aware of them on Greg's forum) possibly at the expense of his products.
As an aside, I wonder how Paulo's is going?
Fixes were only really needed where builders wanted to go outside the standard build parameters. The great thing about Greg is he'd accommodate these people by simming etc. for the new operating conditions or whatever and suggest new parts to enable the amp to work under these conditions.
Strangely, the only real issue was found in really cold conditions where the amp would whine on turn-on, but out excuse here in Australia is it never gets really cold so Greg's test conditions didn't encompass such temperatures. An extra cap fixed this.
He is really helpful even when you ask him anything at all about electronics, including other brands etc. But you can't blame him for not supporting construction of another vendor's boards when they don't work - think about it! Not only that, but the posts he deleted essentially promoted these other boards (that is, made others aware of them on Greg's forum) possibly at the expense of his products.
As an aside, I wonder how Paulo's is going?
Last edited:
A rejection filter for most powerfull AM station in the area
820ohm+18pF. Some people fix it this way
I didn't promote the boards I used. I showed pictures of them so he could see the layout since he suggested the layout may be affecting the oscillation. I posted it in the "Public Domain" thread. I honestly didn't know that he was still trying to sell boards for these after he basically encouraged people to build their own. For people like myself who don't have the ability to etch double sided boards, I have to rely on someone else to make them. It's OK if he doesn't want to help. I just wanted others to know not to expect it. I'm sure there are plenty of guys here who are knowledgeable enough to help me through this.
Blessings, Terry
Blessings, Terry
I had a feeling you might misunderstand me, which is why I added the part in brackets in my last post. I know you didn't promote the boards literally, but people looking at the SKA forum might look at that thread (about the public release of the design) and then your posts would alert then to the existence of cheap eBay PCB's to the detriment of Greg's wares. I'm not judging you, I'm just explaining Greg's reasoning.
Cheers
Stu
Cheers
Stu
I understand your point. My point, is if you are going to release something to the DIY community, you can't then pick and choose who can make them. It's like giving someone a gift and then dictating how they can use it. DX has an issue with Jim's audio too but he has stated many times that he gave him permission to make the boards.
It's all good. I'm still hopeful I'll get this working. It's weird that it worked flawlessly on the bench and then started oscillating once in a case with a earth ground attached. I have installed a loop breaker in the case now so once I get the new devices so I can complete the boards I will give it another try. I have 4 boards populated now. Not sure what I will do with the extra boards.
Blessings, Terry
It's all good. I'm still hopeful I'll get this working. It's weird that it worked flawlessly on the bench and then started oscillating once in a case with a earth ground attached. I have installed a loop breaker in the case now so once I get the new devices so I can complete the boards I will give it another try. I have 4 boards populated now. Not sure what I will do with the extra boards.
Blessings, Terry
Member
Joined 2009
Paid Member
Terry,
The issue is that Jim's Audio is not DIY, it is a commercial venture. Others have reached a license agreement with Jim's Audio and all is good. That's not the case with the SKA GB150 yet - perhaps in the future. In the meantime it's pretty clear on Greg's website that boards are available.
The thing is, if this kind of concern discourages others from releasing schematics for DIY use, we all lose.
I'm confident Greg released his design so we'd take a bash at making it. A DIY pcb design and a group buy might have been more in-line with that.
Anyhow, I don't see you or Greg as bad guys here, you are fellow experimenters and both appreciate the great topology of the GB150 and the good sound - just sharing my perspective because I think we have more to gain from encouraging people like Greg to share his wisdom.
I think you're on the right track with looking at wiring and earthing to solve your troubles. However, this design is sensitive to stray capacitance. The PSRR has a peak at high frequencies that can be tamed with some 'Cdom' on the drivers if the design and parts choices are not as good as Greg's original design (see TGM7 thread).
The issue is that Jim's Audio is not DIY, it is a commercial venture. Others have reached a license agreement with Jim's Audio and all is good. That's not the case with the SKA GB150 yet - perhaps in the future. In the meantime it's pretty clear on Greg's website that boards are available.
The thing is, if this kind of concern discourages others from releasing schematics for DIY use, we all lose.
I'm confident Greg released his design so we'd take a bash at making it. A DIY pcb design and a group buy might have been more in-line with that.
Anyhow, I don't see you or Greg as bad guys here, you are fellow experimenters and both appreciate the great topology of the GB150 and the good sound - just sharing my perspective because I think we have more to gain from encouraging people like Greg to share his wisdom.
I think you're on the right track with looking at wiring and earthing to solve your troubles. However, this design is sensitive to stray capacitance. The PSRR has a peak at high frequencies that can be tamed with some 'Cdom' on the drivers if the design and parts choices are not as good as Greg's original design (see TGM7 thread).
Last edited:
Hi Gareth,
I suppose Jim's audio is a commercial venture but the products they sell are for the benefit of us DIYers. I'm not sure how much money they are making from the boards they sell. They sell them for less than we can have them made, even in a group buy. For Pete's sake, they are $12 a pair.
I don't know them at all except from what I read on their ads and from the small amount of emails I have exchanged with them looking for documentation. They "claim" they were asked to make some boards for this amp. I will not call them liars as I have no evidence to the contrary. I also have no problem with someone trying to find a way to make a living in this world. Both Greg and DX released their designs to public domain. It's not like someone stole their intellectual property, they released it and encouraged folks to build it. I know there are folks who have issues with foreign products, to be honest, almost all of you guys are foreigners to me.
I will check out your GM7 thread but to be honest, I hope to get it working the way it was designed. Paulo seemed to so I have hopes.
Blessings, Terry
I suppose Jim's audio is a commercial venture but the products they sell are for the benefit of us DIYers. I'm not sure how much money they are making from the boards they sell. They sell them for less than we can have them made, even in a group buy. For Pete's sake, they are $12 a pair.
I don't know them at all except from what I read on their ads and from the small amount of emails I have exchanged with them looking for documentation. They "claim" they were asked to make some boards for this amp. I will not call them liars as I have no evidence to the contrary. I also have no problem with someone trying to find a way to make a living in this world. Both Greg and DX released their designs to public domain. It's not like someone stole their intellectual property, they released it and encouraged folks to build it. I know there are folks who have issues with foreign products, to be honest, almost all of you guys are foreigners to me.
I will check out your GM7 thread but to be honest, I hope to get it working the way it was designed. Paulo seemed to so I have hopes.
Blessings, Terry